J. Org. Chem. 1992,57,4546-4550
4546
trations of both 2 and 3 decreased such that the rate dropped slowly. In support of this rationale, the steady second-stage rate was observed from the start in a 1-g hydrolysis of 2 with the soiution phase saturated with 3 initially.
Experimental Section Materials. Acetonitrile (Fisher HPLC-grade) and trifluoroacetic acid (Fisher certified) were used to prepare the HPLC mobile phase. Triton X-100(Roehm and Haas), lipase (Ammo LPL-80 with activity of 889OOO u/g), and aqueous phosphate buffer, prepared from potassium phosphate dibasic (Fisher certified ACS) and concentrated HC1 (Mallinckrodt Analytical Reagent), were used for hydrolysis reaction mixtures. Compounds 2,3, the corresponding racemic ester-acid and diacid, and 4,all described previously,'P2 were available in these laboratories. Kinetics Runs. In heterogeneous enzymatic hydrolyses, solid diester 2 (1.00,0.50, or 0.25 g) was added to 0.1 M pH 7.5 phosphate buffer (30 mL) containing Triton X-100(0.5 mL). The slurries were stirred mechanically and thermostated in a water bath at 40 "C or were agitated using an ultrasonic probe (Heat Systems-Ultrasonics, Inc., Model W-370). The probe was used in a reaction vessel equipped with a circulating water jacket kept at 36 OC to compensate for heat generated by the probe, and the reaction temperature of about 40 OC was monitored using a thermocouple. The hydrolysis was found to proceed the same with mechanical stirring or ulhaonic agitation; mechanical etirring was preferred for better control of the reaction temperature. In preliminary runs, assays of stirred mixtures of the buffer, Triton X-100,and 2 or 3, without enzyme, showed initial supersaturation followed by equilibration within a half hour to fdtrate concentrations (solubilities)of 0.8 m g / d of 2 or 6 mg/mL of 3. In hydrolysis kinetic runsthe lipase (40 mg)was added last to initiate hydrolysis after stirring for a half hour. Aliquota of the sluny, taken periodically while stirring, were analyzed for the diester, ester-acid, and diacid using HPLC; see below. Aliquota of the solution phase, obtained by fdtration through an immersed fine-porosity glass frit, were analyzed similarly. In homogeneous enzymatic hydrolyses, solutions of Triton X-100(0.5 mL) and diester 2 (15mg) in the pH 7.5 buffer (30 mL) were thermostated in the water bath at 40 OC, and again the lipase (20 or 40 mg) was added last to initiate hydrolysis. Some hydrolyses of 2 were performed with 3, the racemic ester-acid, the diacid, or 4 present initially. High-Performance Liquid Chromatography. A 250- X 4.6-mm Partisil 10 ODS column (Whatman) was used for liquid chromatography with an isocratic mobile phase of acetonitrile/0.2 w t % aqueous trifluoroacetic acid 4456 v/v, delivered at a rate of 2 mL/min with the column at ambient temperature (23 "C). Compounds 2, 3 (or the racemate), the diacid, and 4 eluted at retention times of 9,5,3, and 4 min, respectively, and all exhibited the same UV spectrum. Chromatograms were monitored at 292 nm, chosen as a peak wavelength at which Triton X-100and the lipase did not absorb.
Appendix The reversible inhibition of enzyme E by (8-estel-acid product P was modeled assuming rapid equilibria involving unreactive complexes EP1 and EP2 in solution. E + P e EP1 K1 = [EPi]/[E][P] (1) EP1 + P + EP2 K2 = [EP2]/[EP,][P] (2) Hydrolysis of diester substrate S was assumed to occur via complex ES, in rapid equilibrium also. E + S * ES Ks = [ES]/[E][S] (3) The apparent diester reactivity relative to that in the absence of P (the ordinate function in Figure 3) was calculated as [El/[&d]. Assuming [ES] low for the sake of simplicity, EP1 and EP2 were eliminated by substitution from eqs 1and 2 and the enzyme material balance, eq 4, to obtain eq 5.
[%tal1 = [El + [EPlI + PP2I
(4)
[El/[E,tall = 1/[1 + Kl[PI' + K1Kz[Pl21 (5) This function of [PI, evaluated for various values of the ICs, was compared to the relative reactivity data curve, Figure 3. No good fit was obtained assuming a single equilibrium ,eq 1,but calculated curves fitting the data curve were obtained using two equilibria, eqs 1and 2. For example, the curve for K1= K 2 = 0.5 (mg/mL)-' fit, but considerablevariation was found tolerable. The agreement was just as good with K1two or three times larger and Kz moderately smaller, or with Kl two or three times smaller and K z larger. Since [PI exceeded the solubility value of 6 mg/mL during most of the heterogeneous hydrolysis, K values of 0.5 connote [EPl]/[E] and [EPz]/[EPl]ratios greater than 3, eqs 1 and 2.
Acknowledgment. The authors thank Dr. D. L. Hughes for helpful discussions and Mrs. H. R. Graham for typing the manuscript.
Cross Coupling Reactions of 2-(Allyloxy(thio))benzothiazoleswith Organocopper Reagents in Dihydropyranoid Systems. Mechanistic Implications of the Substrate and the Reagent: Regio- and Stereocontrolled Access to Branched-Chain Sugars Serafh Valverde,* Manuel Bernab6, Ana M. Gtimez, and Pilar Puebla Znstituto de Q u h i c a Orgcinica General (CSZC), Juan de la Cierva 3, E-28006 Madrid, Spain Received March 2, 1992
Allylic derivatives are very useful compounds from a synthetic, mechanistic and biochemical point of view.' One of the chemical aspects which has attracted more attention is the regio- and stereochemistry of nucleophilic displacement reactions, and in particular those related to the applications of organometallic reagents in the selective formation of C,C bonds. In general, little regioselectivity has been achieved when working under stoichiometric conditions, and the regioselectivity of the reaction has been shown to depend on such factors as solvent, substrate, reagents, etc. Goering et aLM have achieved a high degree of regiacontrol by using allylic carboxylates in the presence of Grignard reagents and catalytic amounts of cuprous cyanide. The stereochemistry of this reaction with regard to the relative disposition of the metal and the leaving groups is predominantly anti, as in most reactions of allylic electrophiles with electron-rich complexes.2e For these particular substrates these authors have postulated the formation of a u-copper(1II)complex (A) which can either undergo stereospecific reductive elimination to give anti-y alkylation or isomerize to the a-allyl complex (B)(Scheme I). The active species are considered to be RCu(Z)MgBr (1) (a) Magid, R. D. Tetrahedron 1980,36, 1901. (b) Erdik, E. Tetrahedron 1984,40, 641. (2) (a) Underiner, T. L.; Goering, H. L. J. Org. Chem. 1991,56, 2563. (b)Underiner,T. L.; Paisley, S. D.; Schmitter, J.; Leeheski, L.; Goering, H. L. J. Org. Chem. 1989,54,2369. (c) Underiner, T. L.; Goering, H. L. J. O g . Chem. 1989,54,3239. (d) Underiner, T. L.; Goering,H. L. J. Org. Chem. 1988,53,1140. (e) Collman,J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. A.inciples and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987; Chapter 19.
OO22-3263/92/ 1957-4546$03.O0/0 0 1992 American Chemical Society
J. Org. Chem., Vol. 57, No.16, 1992 4547
Notes Scheme I
Table I. Products and Yields Obtained during the Crorr-Coupling Reactionr of Substrates 2,6,9, and 13 entry substrate product yield (96) i 2 3 80 ii iii iv
4
78
I 7
64 50 52
9
8 5 10 11
X
9 13
12 14
xi
13
xii
13
15 16
2 2 6 6 6 9
V
vi vii viii
ix
74
60 60 88 80 76
36
when the cross-coupling reaction was carried out using compound 6, the n-butyl and isopropyl derivatives still reacted in the predicted manner to afford compounds 7 and 8, respectively. However, the organocopper reagent derived from phenyl bromide yielded selectively compound 5 (entry vi). This product still corresponds to a y-substitution, but the stereochemistry, though selective, indicatee an antistereofacial approach of the incoming group. R'- alkyl or phenyl
and is postulated that when Z = CN the 7-cross-coupling prevails, but when Z = R,the allylic rearrangement (U-T rearrangement mechanism) is observed. Mitsunobu et al.3 have also studied the reaction of allylic mesyl derivatives with methyl cuprates, obtaining the methyl derivative corresponding to an a-anti substitution. Though they explain this result as invoking the principle of hard and soft acids and bases (position C-3 is considered softer than position C-1) it is obvious that it could be also explained through Goering's mechanism as well. There are not many examples of the use of allyl ethers in cross-coupling reactions with organocopper reagents. Using an extensive variety of allylic ethers, Normant et al.4 found that the a y ratio was very much dependent on steric effects regardleas of the particular CuX used. In this context, the behavior of a particular type of allylic ethers (2-thio- or 2-oxybenzothiazoles) attracted our a t t e n t i ~ n . ~ Its use with acyclic substrates has permitted workers to secure a certain regiocontrol of the reaction by the choice of appropriate solvents and modification of the addition order of the reagents. We recently showed how this methodology could be adapted to the preparation of pyranoses bearing a methyl branch at C-2 and C-4 with excellent control of regio and stereochemistry? In this context we wondered whether the same methodology could be utilized for introduction of other alkyl groups and for alkyl branching at other sites of the pyranoses. In this report we describe some recent studies in our laboratory dealing with the scope of this protocol.
Results and Discussion Substrate 2 reacted in the expected manner to afford the n-butyl, isopropyl, and phenyl derivatives corresponding to a 7-substitution with syn stereochemistry (3, 4, and 5 respectively) (see Table I). On the other hand, (3)Mitsunobu, 0.; Yoehida, M.; Takiya, M.; Kubo, K.; Mmyama, S.; Satoh, I.; Iwami, H. Chem. Lett. 1989,809. (4)Normant,J. F.; Commercon, A.; Gendreau, Y.; Bourgain, M.; Villierae, J. Bull. SOC.Chim. Fr. 1979,11-309. (5) (a)Cal6,V.; Lbpsz,L.; Carlucci,W. F. J. Chem. Soc., Perkin Trans. I 1983,2953. (b) Cab, V.; Up,L.; Peace,G.;C h o , A. J. Org. Chem. 1982,47, 4482. (6) Valverde, S.;Bernabe, M.; Garcia-Ochoa,5.;Mmez, A. M. J. Org. Chem. 1990,55,2294.
d
BnO R' 2
6
R2
BTZ-S H H BTZ-0
R2
R'
9 BTZ-S H 13 H BTZ-0
R'
R2
10
n-Bu
H
H
11
i-Pr
H
Ph
H
12
Ph
H
7
H
n-Bu
14
H
n-Bu
8
H
i-Pr
15
H
i-Pr
R'
R2
3
n-Bu
H
4
i-Pr
5
Ctm
C.l
TBDMSO
16
Alkylation of the thiobenzothiazole 9 and oxybenzothiazole 13 took place yielding the respective n-butyl and isopropyl derivatives, according to the expected regio- and stereochemistry (10,11,14,15). When phenylmagnesium bromide in presence of copper iodide was used in the cross-coupling reaction of 9, compound 12 wm exclusively produced in 88% yield, showing that the reaction follows the same pattern as the alkylation process. However, when the oxybenzothiazole derivative 13 was allowed to react under similar conditions compound 16 was obtained in 36% yield (entry xii). Arylation has thus occurred with stereo- and regicontrol, but no allylic transposition was observed. As an additional example of this cross-coupling reaction, compound 23 waa next investigated. Treatment of 23 (Scheme 11) with methyl cuprate or phenyl cuprate afforded the expected products 24 and 25, respectively.
4648 J. Ore. Chem., Vol. 57, No.16, 1992
Notes
Table 11. 'H NMR Significant Spectral Parameters for Compounds 2-5,7,8,10-12.14-16, H-1 H-2 H-3 H-4 H-5 H-6 H-6' Jiz JlS Jl4 Jis J2a J24
5 2 6 Js4
536 J46
2
3
5.123 5.897 6.229 5.909 4.227 3.717 3.732 2.8 -1.5 1.6 -0.7 10.3 -1.8
4.749 2.024 5.810 5.675 4.358 3.513 3.566 0.9 -1.2 -0.3 -0.5 5.0
-1.6 2.7 ca. 0 10.4 1.8 -2.3 CaO 1.6 9.5 4.6 4.2 6.6 2.5 -11.1 -10.0
Jsa Jm' Jee' a Solvent: C &.
4"
5
4.810 4.861 1.995 3.343 5.630 5.864 5.828 5.920 4.478 4.450 3.444 3.679 3.582 3.716 1.1 0.9 -1.2 -1.2 -0.3 -0.3 -0.5 -0.4 4.8 4.9 -1.9 -1.8 3.1 3.6 10.6 10.4 -2.3 -2.3 1.8 1.8 4.3 5.7 5.8 6.1 -9.9 -9.6
7'
4.850 2.334 5.571 5.738 4.485 3.500 3.622 4.4 -1.1
-0.3 -0.7 2.2 2.4 3.9 10.3 -2.2 2.1 5.5 6.0 -9.7
8
10
11
12"
14
4.987 2.059 5.818 5.699 ca 4.33 ca. 3.53 ca. 3.56 4.1
4.825 5.718 6.131 2.020 4.036 3.646 3.726 2.7
4.814 6.039 2.010 4.059 3.717 3.781 2.9
-1.3 ca. 0 ca. 0 1.8
-1.1
-1.1
0.3 -0.6 10.1 -1.2 ca. 0 5.9 ca. 0 3.1 6.5 7.0 -10.3
C a O
4.848 5.784 5.854 3.162 4.610 3.551 3.639 2.9 -1.0 ca. 0 -0.6 9.9 -1.4 ca. 0 5.7 ca. 0 3.6 6.7 6.5 -10.3
4.847 5.732 5.920 2.180 3.602 3.743 3.813 3.0 -1.4 1.5 -0.3 10.1 -2.8 ca. 0 1.9 ca. 0 10.0 5.7 2.5 -11.3
-2.7 4.3 10.5 -2.1 2.0 ndb ndb ndb
5.866
-0.9 10.2 -1.3 ca. 0 5.5 -0.2 3.9 6.2 7.4 -10.4
15
and 23-25
16
4.757 4.613 5.719 3.219 5.818 5.725 2.096 5.842 ca. 3.62 4.467 ndb 3.671 ndb 3.744 3.0 1.2 -1.3 -1.2 1.6 ca. 0 ca. 0 -0.4 10.2 4.8 -1.7 -3.0 ca. 0 3.1 2.0 10.4 -2.3 ca. 0 10.0 1.7 ndb 5.5 ndb 5.3 ndb -10.4
23"
24"
26"
6.449 5.236 5.553 3.851 3.692 3.345 3.520 2.7 -2.0 2.5 -0.5 10.1 -2.6
6.132 4.510 2.219 3.909 3.741 3.676 3.940 5.9 -1.6 ca. 0 ca. 0 5.8 ca. 0 ca. 0 6.4
6.088 4.304 3.175 3.791 3.520 3.378 3.585 6.0 -1.3
ca. 0
ca. 0
9.9 10.3 5.4 -10.7
9.7 10.2 5.3 -10.6
ca. 0
1.7 -0.3 9.0 10.6 5.0
-10.5
ca. 0 ca. 0
5.9 ca. 0 ca. 0 6.1
nd = not determined.
Clearly alkylation always followed the same pattern, regardless of the substrate, affording the alkyl derivatives that correspond to a y-syn substitution. In all these cases alkylation could take place through the mechanism depicted in C (Scheme I). On the contrary, arylation has followed different patterns for substrates 2,9, and 23 (on one hand) and 6 and 13 (on the other). In the first case, when the benzothiazole group is on the &face (e.g., opposite to the substituent at the anomeric center) the reaction proceeds in a y-syn manner, affording compounds 5 and 12 (entries iii and ix) as in the case of alkylation. However, substrates 6 and 13 with the benzothiazole group at the a-face afforded compounds 5 and 16, respectively (entries vi and xii), which correspond to anti substitutions. In order to explain the specific formation of a- or y-anti products we postulate that in the last two cases the rel action proceeds through the formation of a ~ d ycomplex B as reaction intermediate, following the suggestion of Goering et To support this interpretation we carried out the reaction of 21 with methylmagnesium iodide and CUI (Scheme 11). As previously reported in the literature,' for methylations of unsaturated dihydropyrans (compounds 18 and 20) and according to the mechanism previously proposedk for cross-couplingallylic carboxylates with alkyl(sp8)copper reagents, this reaction should take place through complex A affording the methyl derivative correspondingto a y-anti substitution. In fact, compound 21 afforded 22. To explain this result, in terms of the mechanism proposed by Goering et d.,2"d it is necessary that the u-aUyl-copper(DI)complex A isomerizes to the r-allyl complex B prior to methylation. Since the only isolated product corresponds to an a-anti substitution, the dihydropyran substrate must be strongly biased in favor of substitution taking place at C-2 (pyrano86 numbering). An analogous explanation would be valid for the reaction of phenyl(sp2)copperreagents with substrates 6 and 13 which also led to C-2 anti-substituted producta, independently of whether the starting material was a 2,3 or a 3,4-uneaturated pyranose. Finally, in the case of 23, where the benzothiazole group is in the anomeric position, a y-syn substitution was observed both for (7) (a) Danishefsky, S. J.; Armistead, S. M.; Wincott, F. E.; Selnick, H.G.; Hungate, R. J. Am. Chem. Soc. 1989,111,2967. (b)Chapleur, Y.; Crapsas, Y . Carbohydr. Res. 1985,141,163. (c) Boquel, P.; Chapleur, Y .
Tetrahedron Lett. 1990, 31, 1869.
Scheme I1
17
18
10
20
21
22
R
23
24 R=CH3 25 R-Ph
Pv-Pivaloyl
the alkylation and arylation reactions. Possibly, in the case of the phenyl(sp2)copperreagents, the formation of complex C has such electronic or steric requirementa that cannot be provided by substratea 6 and 13 (due to the proximity of the substituent at the anomeric center). Consequently, the reaction occurs through the alternative pathway represented by complex B. Compound 23 where no interference from the anomeric substituent was possible reacted with phenylcopper reagenta yielding the y-syn product. None of the substrates men-
J. Org. Chem., Vol. 57, No. 16, 1992 4649
Notes
tioned above (2,6,9,13, and 23) afforded alkyl derivativea when the crm-coupling reaction was attempted using tert-butylmagnesium bromide. Structural Features. The structures of all the compounds obtained were assigned through the analyses of their 'H NMR data (see Table 11). The 300-MHz 'H NMR spectra of all derivatives were analyzed iteratively, and the beat computed data of chemical ehifts and coupling constants are given in Table II. The sign of the coupling constants were assessed from the literatureeor obtained from the calculations. The value of Jl,z(ca. 4 Hz in compounds 7 and 8 and c a 1 Hz in derivatives 3, 4, 5, and 16) indicates equatorial-qwi-axial and equatorial-quasi-equatorialgeometries for H-1 and H-2, respectively. On the other hand, the magnitude of the coupling constants J4,6(3-4 Hz in compounds 10-12 and ca. 10 Hz in derivatives 14 and 15) suggest quasi-equatorial-axial and quasi-axial-axial arrangements of H-4 and H-5, respectively, and are in accordance with the values calculated from Altona's equation? The negligible valuea of homdylic couplings (J1,4) and those of J3,4(ca. 5.5) are in support of H-4 being quasi-equatorial in compounds 10-12, while the magnitudes of JSC(ca. 2 Hz) and J1,4 (ca 1.5 Hz) suggest a quasi-axial geometry for H-4 in compounds 14 and 15. Concerning the stereochemistry of the bicyclic derivatives 24 and 25, the values of J3,4(6.14.4 Hz) are in agreement with those calculated according to Altona's equation for angles in the neighborhood of 40°, which indicates quasi-axial arrangements of the substituents at (2-3. NOE experiments were also performed on selected compounds, leading to unambiguous location of the double bond and providing additional support to the structures deduced from coupling constants. Thus, irradiation of H-5 produced around a 6% increase in the intensity of H-4 in 10 and 11, and no noticeable variation in analogous proton of 14, while irradiation of H-2 led to incrementa of 4% in the intensity of H-1 in compounds 5 and 16 and 6% in that of 7. Finally, irradiation of H-3 in compound 26 induced a 5% increase in the intensity of H-4. As expected, all the data are in accordance with predominant OH5conformations for compounds 2 and 10-15 and OH1 conformations for 3-8 and 16. From the results obtained it could be concluded that a complete control of the regie and stereoselectivity can be achieved in the formation of C-C bonds in dihydropyrans by the w of the oxy(thio)benzothiazolylgroup. C-Alkyl or C-aryl derivatives of predefined stereochemistry can be obtained at positions C-2, C-3, and C 4 of the pyranosidic ring. In the alkylation process the regie and stereochemical outcome is dictated by coordination effects, in such a manner that a complex is formed between the copper species and the substrate. However, in the arylation reaction the substrate has a decisive role on the control of the process. Experimental Section For general experimental detaih see ref 6. Starting Materials. The known diols, ethyl 2,3-dideoxy-a~-erythro-hex-2-enopyranoaide (la)and methyl 3,4-dideoxy-ape&wo-h-3-enopyranoside (lb)were prepared from ~glucoee, as previously described? 4,6-O-Isopropylidene-~-glucal was prepared from commercial glucal.'O Ethyl 2,3-dideoxy-6-0(8)Achmatowia, O., Jr.; Bauazek, A.; Chmieleweki,M.; Zamojeki, A.; Lobodzinski, W. Carbohydrate Res. 1974, 36, 13 and references cited therein. (9) Haasnot, C. A. G.; de Leeuw, F. A. A; Albne, C. Tetrahedron 1980, 36,2783.
benzyl-a-~-erythro-hex-2-enopyran~ide (IC)," methyl 3,4-dideoxy-6-0-(tert-butyldimethylsilyl)-a-~-erythro-hex-3-enopyranoside (la),@ and substrates 2,12 9: and lSSwere identified either by identity with authentic samplee or comparison with previously published spectroscopic information. All known compounds gave satisfactory phyeical and spectral data coneistent with their structures. Product 312was ale0 identified through ita spectroscopic data General Procedure for the Synthesis of (Allylthio)was benzothiazoles 2,9, and 23. The allylic alcohol (5 "01) diesolved in toluene (20 mL) together with PPh3 (6 "01) and 2-mercaptohmthiazo1(5.5 "01). While the temperature was maintained below 5 "C, diethyl azodicarboxylate (5.5 "01) diaeolved in toluene was added dropwise and with stirring. Once the allylic alcohol was consumed the suspension was filtered and washed with toluene, and the solution was concentrated in vacuo. The residue was then directly applied on a silica gel column. General Procedure for the Synthesis of the (Allyloxy)benzothiazoles 6 and 13. To the allylic alcohol (7 mmol) dissolved in dry ether (60 mL) was added metallic potassium (10.5 "01) in small pieces at 0 "C and with stirring. Once the reaction w a e ~ p l e t e , w o r (10.5 o ~ "01) dieeolved in ether (5 mL) was added, and the reaction mixture was left at room temperature for 24 h. Methanol was then added until complete disappearanceof excess metal. The reaction mixturewas washed with water, dried, and evaporated. The oily residue was chromatographed ( h e m t h y 1 acetate (82)) to give the product. Ethyl 2,3-Mdeoxy-6-O-benzyl-k0-(2-benzothiazolyl)-a~erytlvo-hex-2-emopyranoside(6). This compound, a thick Oil, WBB SYllthdtd from IC (70%): ["ID 115.0" (C 0.71, CHClJ; 'H NMR (300MHz) 6 1.26 (3 H, t, J 6.3 Hz, OCHaCHg), 3.60 and 3.90 (2 H, 2 dq, J#em = 9.8 Hz, OCHzCHg), 3.72 (2 H, 9, Je,v = 12.0 Hz,H a , H-6'),4.15 (1 H, m, H-5),4.56 (2 H, q, J 12.0 Hz,OCHQh), 5.12 (1 H, d, J19= 2.6 Hz,H-l), 5.91 (1 K d d , J21 = 2.6 Hz,J = 10.0 Hz,H-2),5.90 (1H, m (overlapping with H-2), H-4), 6.22 H, d, Jg,2 = 10.0 Hz, H-3), plus signale due to the benzothiazolyl group and aromatic benzyl group. Anal. Calcd for CBHBNO4S: C, 66.48, H, 5.83; N, 3.52; S, 8.0. Found C, 66.75; H, 5.75; N, 3.70; S, 8.30. General Procedure for the Reaction of (Allyloxy(thio))benzothiazole Derivatives with Organocopper Reagents. The Grignard reagent was prepared in anhydrous ether (10 mL) from Mg (4.0 "01) and RBr (R = n-Bu, LPr, t-Bu, Ph) (4.0 "01). This solution was cooled to -30 "C and CUI (1.5 "01) was added in one portion under argon. Stirring was continued during 30 min at -30 OC, and then a solution (15 mL) of the in ether was added. In order to obtain substrate (1.0 "01) acceptable yields and to ensure high regio- and stereoselectivity, a good stirring of the reaction mixture was critical. The mixture was allowed to warm slowly to mom temperature and stirred for 6 h. The reaction was diluted with ether and treated with concentrated aquoue NH4Cl and a few drops of NH40H, while vigorously stirring. The ethereal phaee was separated and dried over anhydrous sodium sulfate and concentrated in vacuo. Flash chromatography of the crude midue (hesaneethyl acetate (99)) yielded the product. Ethyl 2,3,4-Trideoxy-2-C-isopropyl-6-O-benzyl-a-Dtluwo-hex-3+nopyranoside(4). This compound was prepared from 2 and isopropylmagneaium bromide as a syrup (78%): [aID 50.9" (c 0.21, CHClJ. Anal. Calcd for ClaHzaOs: C, 74.46; H, 9.02. Found C, 74.40; H, 9.05. Ethyl 2$,4-Trideoxy-2-C-phenyl-6-0 -benzyl-a-D-t h m hex-t-enopyranoside (5). Prepared from 2 or 6 and phenylmagnesium bromide as a syrup (64% and 52%, respectively): [& 148.6" (c 0.95, CHClJ. Anal. Calcd for C21HaO3: C, 77.74; H, 7.46. Found C, 77.92; H, 7.20. Ethyl 2,3,4-Trideoxy-2-C -n-butyl-6-0 -benzyl-a-Derytliro-her-t-enopyranoside(7). Thie compound was synthesized from 6 and n-butylmagnesium bromide as a colorlees oil (74%): [a]D -23.0" (c 0.51, CHCl3). And. Calcd for C19H2s03:
-
3
-
(10) Fraeer-Reid, B.; Walker, D. L.;Tam,S. Y.K.; Holder, N. L.Can. J. Chem. 1979,51,3950. (11) Valverde, 5.;Garcia-Ochoa, S.; Martin-Lomas, M. J. Chem. Soc., Chem. Commun. 1987,1714. (12) Neirabeyeh, M. A.; Rollin, P. J. Corbohydr. Chem. 1990,9,471.
4650
J. Org. Chem. 1992,57,4550-4552
74.96; H, 9.27. Found C, 75.13; H, 9.33. Ethyl 2,3,4-Trideoxy-2-C-isopropyl-60 -benzyl-a-Derytbro-hex-knopyranoside(8). This product waa prepared from 6 and isopropylmagnesium bromide as a syrup (50%): [U]D -19.6O (c 0.7, CHC18). Anal. Calcd for CleHZeO8: C, 74.45; H, 9.02. Found C, 74.41; H, 8.87. Methyl 2,3,4-Trideoxy-4-C-n -butyl-6-O-(tert-butyldimethylsily1)-a-D-th-hex-2-enopyranoside (10). Prepared from 9 and n-butyImagnesium bromide aa a colorleas oil (60%): [ a ]-71.6' ~ (C 0.72, CHCld. Anal. calcd for Cl7H&& c,61-92; H, 10.9. Found C, 64.65; H, 10.65. Methyl 2,3,kTrideoxy-4-C-ispropyl-60 -(tert -butyldimethylsilyl)-a-~-tlureo-hex-2-enopyranoside (11). Prepared from 9 and isopropylmagnesium bromide as a syrup (60%): [a]D -85.0° (c 1.4; CHCla). Anal. Calcd for C1a9203Si: C, 63.95; H, 10.74. Found C, 64.13; H, 10.60. Methyl 2,3,4-Trideoxy-4-C-pheny1-6-0-( tert -butyldimethylsilyl)-a-~-tlureo-hex-2-enopyranoside (12). Prepared from 9 and phenylmegnesium bromide as a syrup (88%): ["ID -104.2O (c 0.8; CHC13). Anal. Calcd for C1$Im03Si: C, 68.22; H, 9.04. Found C, 68.41; H, 8.82. Methyl 2,3,4-Trideoxy-4-C-n -butyl-6-0-(tert -butyldimethylrilyl)s-D--hex-2-enopyraaoside (14). Prepared from 13 and n-butylmagneeium bromide as a colorless oil (80%): Cdcd for C17HM03Si:c,61-92 ["ID 33.2' (C 0.42, CHCls). H, 10.9. Found C, 65.12; H, 10.72. Methyl 28,4-Trideoxy-4-C-isopropyl-6-0-(tert -butyldimethylsilyl)s-De-hex-~opyranoside(15). Prepared from 13 and i s o p r o p y l " bromide as a syrup (76%): [a]D 77.4' (c 0.44,CH2Cl&. Anal. Cdcd for C16H3203Si:c, 63.95; H, 10.74. Found: C, 64.15; H, 10.50. Methyl 2,3,4-Trideoxy-2-C-phenyl-6-0 -(tert-butyldimethylsilyl)-a-D-th-hex-3-enopyranoside (16). Prepared from 13 and phenylmagnesium bromide as a syrup (36%). Starting material (50%) was recovered unchanged ["ID 77.3O (c 0.7; CHC13). Anal. Calcd for Cl&Im03Si: C, 68.22; H, 9.04. Found C, 68.33; H, 8.77. Methyl 2 ~ ~ - T r i d ~ ~ 2 - C - m e t h y -benzyl-a-P 1-6-0 t h hex-3-enopyranoside (22). Methyl iodide (568 mg, 4.0 mmol) in anhydrous was reacted with metallic Mg (97 mg, 4.0 "01) ether. This solution was cooled at -30 OC, and CUI(286mg,1.5 mmol) was added in one portion under argon. Stirring was continued during 30 min at -30 OC, and a solution of 2113 (368 mg, 1.0 "01) in ether (15 mL) was added. This mixture was allowed to warm slowly to room temperature and stirred for 6 h. The reaction waa diluted with ether and treated with concentrated aqueous NH,Cl and a few drops of NH,OH while vigorously stirring. The ethereal layer was separated and dried over anhydrous sodium sulfate and concentrated in vacuo. The residue was rebemylated (BQO/Py) and chromatographed on silica gel to yield the product, a thick oil (80mg, 30%). 2-Benzot hiazolyl 2,3-Dideoxy-4,6-0 4sopropylidene-1thio-a-Deryt~-hex-e.-2enopyranoside (23). A syrupy product and 2prepared from 4,6-di-O-isopropylidene-~-glucal~~ mercaptobenzothiazole using the general conditions described a e ~310.9O : * ~ (c 0.7, CHC13). above for ( a l l y l t h i ~ ) b e ~ ~ ~ t h i["ID Anal. Calcd for C1&,N0& C, 57.29;H, 5.11; N, 4.18; S, 19.12. Found C, 57.40; H, 5.11; N, 4.15; S, 19.44. 1,5-Anhydro-4,6-0 -isopropylidene-2,3-dideoxy-3C methyl-D-ribo-hex-1-enitol(24). Prepared from 23 and MeMgBr as a syrup: [a]D 160.9' (c 1.4, CHC13). Anal. Calcd for C&1@3: C, 65.19; H,8.75. Found: C, 65.10; H,8.90. l,S-Anhydro-4,6- 0 -isopropylidene-2,3-dideoxy-3-C phenyl-~ribo-hex-l-enitol(28). Prepad from 23 and PhMgBr aa a syrup: [& 244.3O (c 0.86, CHCl3). Anal. Calcd for C1,5HlBO3: C, 73.15; H, 7.37. Found: C, 73.31; H, 7.60.
-
-
Acknowledgment. Financial support from DGICYT (PB90-0078)is gratefully acknowledged. (13) Holder, N. L.; Fraeer-Reid, B. Can. J. Chem. 1973, 51, 3357. (14) Substitutions with subsequent allylic rearrangement have been previously reported to take place during the Mitsunobu reaction with dihydropyran derivativea (Dyong, I.; Weigand, J.; Thiem, J. Liebigs Ann. Chem. 1986,577). A minor component, possibly a 3-benzothiazolyl glycal derivative, was also detectad.
Registry No. la, 23339-15-3; lb, 51385-38-7; IC,80516-25-2; ld, 124944-63-4; 2,130619-67-7; 8, 130641-42-8;4,142041-72-3; 5, 142041-73-4;6,142041-74-5; 7,142041-756; 8,142041-76-7;9, 124944-67-8; 10, 142041-77-8; 11, 142041-78-9; 12, 142041-79-0; 13,124944-66-714,142041-80-3;15,142041-814,16,142041-82-5; 21,34254-52-9;22,142041-83-6; 23,142041-84-7;24,142041-85-8 25, 142041-86-9; BTZC1, 615-20-3; BTZSH, 149-30-4; 4 , 6 - ~ isopropylidene-D-glucal, 51450-36-3.
Nickel-Catalyzed Geminal Dimethylation of Allylic Cyclic Dithioketals. A Convenient Procedure To Form a tert-Butyl Substituent at the Olefinic Carbon Atom Tien-Min Yuan and Tien-Yau Luh+ Department of Chemistry, Nationul Taiwan University, Taipei, Taiwan 106, Republic of China Received January 2, 1992
Much effort has been devoted to the introduction of a tert-butyl group or a quaternary carbon to olefinic carbon atom(s) with the aim of synthesizing crowded olefins.' Most of the procedures that have been developed w e reagents containing a tert-butyl group. Although Tebbelike reagents are effective for converting a carbonyl group into a gem-dimethyl substituent, their application to allylic carbonyl substrates is limited by poor regioselectivity.2 We recently reported a series of nickel-catalyzed crosscoupling reactions of benzylic and allylic dithioacetalswith Grignard In the presence of NiCl,(dppe), cinnamaldehyde dithioacetals 1 (R = Ar,R' = H) react with MeMgI to give the geminally dimethylated produde (eq 1,R = h,R' = H).4 We have extended this reaction to geminal dimethylation of allylic dithioketals and now report a facile procedure for the regioselective preparation of tert-butyl-substituted olefins 2 (eq 1, R' # H).
1
2
Dithioketals 1 were prepared according to a modified literature procedure? Treatment of 1 with 4 equiv of MeMgI in the presence of 5 mol % of NiC12(dppe) in refluxing ether-THF for 10 h afforded gem-dimethyl (1) For recent leading references, see: (a) Mulzer, J.; Lammer, 0. Angew. Chem.,Znt. Ed. Engl. 1983,22,628. (b)Russell, G. A; Teshtoueb, H.; Ngoviwatchai,P. J. Am. Chem. Soc. 1984,106,4622. (c) Eisch, J. J.; Behrooz, M.; Ode, J. E. Tetrahedron Lett. 1984,25,4851. (d) Kirkuchi, 0.; Yoshida, H. Bull. Chem. SOC. Jpn. 1988,58, 131. (e) Ager, D. J. J . Chem. Soc., Perkin !IYans. 1 1986,183. (0 Aguero, A.; Krees, J.; Osbom, J. A. J. Chem. SOC., Chem. Commun. 1986,531. (9) Smegal, J. A; Meier, I. K.; Schwartz, J. J. Am. Chem. SOC.1986,108,1322. (h) Palomo, C.; Aizpurua, J. M.; Garcia, J. M.; Ganboa, I.; Coeeio, F. P.; Lecea, B.; Lopez, C. J. Org. Chem. 1990,55,2498. (2) (a) Poener, G. H.; Brunelle, D. J. Tetrahedron Lett. 1972,239. (b) Meisters, A.; Mole, T. J. Chem. Soc., Chem. Commun. 1972, 595. (c) Posner, G. H.; Brunelle,D. J. Tetrahedron Lett. 1973,935. (d) Meieters, A.; Mole, T. A u t . J. Chem. 1974,27,1655. (e) Reetz,M. T.; Weatermann, J.; Steinbach, R. Angew. Chem., Znt. Ed. Engl. 1980,19,900. (0 Reetz, M. T.; Westermann, J.; Steinbach, R. J. Chem. SOC.,Chem. Commun. 1981,237. (g) Brown-Weneley, K. A.; Buchwald, S. L;Cannizzo, L.; Ho, C. 5.;Meinhardt, D.; Stille, J. R.; Straw, D.; Grubbs, R. H. Pure Appl. Chem. 1988,55,1733. (h) Pine, S. H.;Pettit, R. J.; Geib, G. D.; Cruz, S. G.; Gallego, C. H.;Tijerina, T.; Pine, R. D. J. Org. Chem. 1985,50,1212. (i) Reetz, M. T.; Westermann, J.; Kyung, 5.-H. Chem. Ber. 1986, 118, 1150. (3) For a review, see: Luh, T.-Y. Aec. Chem. Res. 1991,24,257. (4) (a) Yang, P.-F.;'Ni,Z.-J.; Luh,T.-Y.J . Org. Chem. 1989,54,2261.
(b) Tzeng, Y.-L.; Yang, P.-F.; Mei, N.-W.;Yuan, T.-M.;Yu, C. C.; Luh, T.-Y. J. Org. Chem. 1991,56, 5289. (5) Ni, Z.J.; Mei, N.-W.;Shi, X.; Wang, M. C.; Tzeng, Y.-L.;Luh, T.-Y. J. Org. Chem. 1991,56,4035. (6) Williams, J. R.; Sarkieian, G. M. Synthesis 1974, 32.
0022-3263/92/1957-4550$03.00/00 - 1992 American Chemical Society